This project is designed to generate a number of distinctive animations on a grid of LEDs via OPC, to a FadeCandy board.
If you're not already familiar, go check out https://github.com/scanlime/fadecandy for more background on this project.
The code has been run on OSX with Python 2.7, and there are a few extras you'll need before you can start. Typically you'd run something like:
# pip install -r requirements.txt
as the super user to install. Or maybe - if you prefer - you'd install in a virtualenv. On the off-chance that the installation gets a hiccup:
-
numpy is pretty well documented and you should be able to figure it out. Some distros have a natively installable package to use instead of the pip equivalent. If you ido install bug get crappy performance, then it may be that there are some ancillary packages missing which numpy relies on for optimum performance. Again, the numpy docs are pretty good.
-
Either Pillow or it's predecessor PIL should work just fine, just make sure that you have one or the other installed. Probably not both though.
To get started, update dpyinfo.py to suit your display configuration, start the OPC service and connect your array. Then run:
% python art.py
You can also pick a specific subset to render by listing the particular modules in the art directory, for example:
% python art.py rain.py fortune.py
If you don't have a display handy, then that's okay. The default output is to the controlling tty, so even if you're just checking this stuff out or on the road, then you can still get an idea of the animations. The code will auto-detect whether the terminal supports color and use it if it can.
Be aware that the down-sampled image is super lossy, since it's reducing 24 bit color to a handful of ascii-art characters... although if your tty has a 256 color mode, then things will look quite a bit better. Also, don't be too adventurous in trying to render to areas larger than the tty, as curses will choke. In this case, you'll see a message saying that the terminal is too small. The terminal needs to be three lines taller than the matrix in order to work.
The config.py file contains tunable parameters relating to the display,
such as geometry, as well as configuration options specific to the
individual art files.
Most textual logging is delivered to art.log. If something doesn't look right, then check in this file for exceptions and other diagnostics.
It's also possible to limit number of cycles that art.py runs through,
adjust the duration that each art is displayed before moving on to the next,
or to switch on profiling. Run with --help for more information.
The directories are
. application directory
./art contains classes for each of the animations.
You'll notice a template.py file for the bare
bones
./art/utils helper modules shared between animations
./art/baseclasses base classes that animations may derive from
./assets files that are used by the animations
./config.py various configuration options
./opc an extended interface for using OPC via python
./opc/utils various utilities used by the framework
Check out the existing animations for use as examples, and please commit back code that you create!
If you don't have a physical display, or a suitable terminal, then it's still possbile to see what's going on by looking at the animations as a web page. For this, you'll need to run the bundled Flask app like this:
% python webrop.py
You should then be able to go visit the web page at http://localhost:5000
If course, this is using the bundled development web server that comes with Flask. If you want to serve volumes of this stuff, then it'll be better to run under a production web server.
A random assortment of images documenting the development of the hardware can be found on Imgur.
The opc code is taken from https://github.com/scanlime/fadecandy.
Hopefully attriibutions are correct, if you see one missing, then please let me know and I'll fix.